ArticlePDF Available

Abstract and Figures

An analytic cognitive style denotes a propensity to set aside highly salient intuitions when engaging in problem solving. We assess the hypothesis that an analytic cognitive style is associated with a history of questioning, altering, and rejecting (i.e., unbelieving) supernatural claims, both religious and paranormal. In two studies, we examined associations of God beliefs, religious engagement (attendance at religious services, praying, etc.), conventional religious beliefs (heaven, miracles, etc.) and paranormal beliefs (extrasensory perception, levitation, etc.) with performance measures of cognitive ability and analytic cognitive style. An analytic cognitive style negatively predicted both religious and paranormal beliefs when controlling for cognitive ability as well as religious engagement, sex, age, political ideology, and education. Participants more willing to engage in analytic reasoning were less likely to endorse supernatural beliefs. Further, an association between analytic cognitive style and religious engagement was mediated by religious beliefs, suggesting that an analytic cognitive style negatively affects religious engagement via lower acceptance of conventional religious beliefs. Results for types of God belief indicate that the association between an analytic cognitive style and God beliefs is more nuanced than mere acceptance and rejection, but also includes adopting less conventional God beliefs, such as Pantheism or Deism. Our data are consistent with the idea that two people who share the same cognitive ability, education, political ideology, sex, age and level of religious engagement can acquire very different sets of beliefs about the world if they differ in their propensity to think analytically.
Content may be subject to copyright.
Analytic cognitive style predicts religious and paranormal belief
Gordon Pennycook
, James Allan Cheyne, Paul Seli, Derek J. Koehler, Jonathan A. Fugelsang
Department of Psychology, University of Waterloo, Canada
article info
Article history:
Received 21 September 2011
Revised 6 March 2012
Accepted 8 March 2012
Available online 4 April 2012
Keywords:
Religiosity
Paranormal beliefs
Supernatural beliefs
Cognitive ability
Cognitive style
Intuition
Dual-process theories
abstract
An analytic cognitive style denotes a propensity to set aside highly salient intuitions when
engaging in problem solving. We assess the hypothesis that an analytic cognitive style is
associated with a history of questioning, altering, and rejecting (i.e., unbelieving) supernat-
ural claims, both religious and paranormal. In two studies, we examined associations of
God beliefs, religious engagement (attendance at religious services, praying, etc.), conven-
tional religious beliefs (heaven, miracles, etc.) and paranormal beliefs (extrasensory per-
ception, levitation, etc.) with performance measures of cognitive ability and analytic
cognitive style. An analytic cognitive style negatively predicted both religious and paranor-
mal beliefs when controlling for cognitive ability as well as religious engagement, sex, age,
political ideology, and education. Participants more willing to engage in analytic reasoning
were less likely to endorse supernatural beliefs. Further, an association between analytic
cognitive style and religious engagement was mediated by religious beliefs, suggesting that
an analytic cognitive style negatively affects religious engagement via lower acceptance of
conventional religious beliefs. Results for types of God belief indicate that the association
between an analytic cognitive style and God beliefs is more nuanced than mere acceptance
and rejection, but also includes adopting less conventional God beliefs, such as Pantheism
or Deism. Our data are consistent with the idea that two people who share the same cog-
nitive ability, education, political ideology, sex, age and level of religious engagement can
acquire very different sets of beliefs about the world if they differ in their propensity to
think analytically.
Ó2012 Elsevier B.V. All rights reserved.
‘‘It is the heart which perceives God and not the reason.
That is what faith is: God perceived by the heart, not by
the reason.’’
Blaise Pascal (1623–1662)
1. Introduction
Belief in beings, forces, or powers that are non-material,
or otherwise with features outside the daily experience of
most people, is found in all human cultures. Zuckerman
(2007), for example, has estimated that roughly 90% of
the world’s population believes in some form of deity. The-
istic beliefs are universally accompanied by a variety of
additional specific supernatural beliefs as well as various
religious rituals and practices. Other kinds of supernatural
belief, commonly referred to as ‘‘paranormal’’, are also
common. For example, more than 40% of Americans be-
lieve in ghosts, spiritual healing, and extra sensory percep-
tion (National Science Foundation, 2002; Rice, 2003).
Recent accounts of the origins of religiosity and religion
have emphasized the intuitive and sometimes ‘‘minimally
counterintuitive’’ nature of religious beliefs, generally
making the case that such beliefs are a natural by-product
of normal human cognition (Atran, 2002; Barrett, 2000;
Boyer, 1994; Frey, 2009; Guthrie, 1993; Lawson, 2000;
Pyysiäinen, 2001). On the other hand, increasing numbers
of individuals in modern societies find religious and para-
0010-0277/$ - see front matter Ó2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cognition.2012.03.003
Corresponding author. Address: Department of Psychology, Univer-
sity of Waterloo, 200 University Avenue West, Waterloo, ON, Canada N2L
3G1.
E-mail address: gpennyco@uwaterloo.ca (G. Pennycook).
Cognition 123 (2012) 335–346
Contents lists available at SciVerse ScienceDirect
Cognition
journal homepage: www.elsevier.com/locate/COGNIT
normal beliefs not only counterintuitive, but quite incred-
ible (Beit-Hallahmi, 2006; Zuckerman, 2007). Why do
some people hold very strong religious beliefs while others
are quite dubious of them? Answers to this question will
almost certainly involve many factors at many levels
including affective, experiential, family, institutional,
developmental, and cultural variables, among others. The
rather ambiguous connection between intuition and the
supernatural does, however, link cognitive theories of reli-
giosity with decades of decision-making literature that
suggests intuition plays a fundamental role in reasoning
processes.
1.1. Intuitive and analytic cognitive styles
Considerable research in recent decades has focused on
two contrasting styles of problem-solving and decision-
making, often formalized as distinct reasoning types or
systems (e.g., Epstein, 1994; Evans, 2008; Kahneman,
2003; Sloman, 1996; Stanovich, 2009). The first, sometimes
referred to as Type 1 processing, is characterized as intui-
tive, fast, unconscious, associative, and heuristic. Alterna-
tively, problem solving and decision-making sometimes
proceeds in a more analytic manner, sometimes called
Type 2 processing, which tends to be more time-consum-
ing, deliberative, and effortful. An analytic cognitive style
will typically involve a broader assessment of problem ele-
ments as well as an examination and critical evaluation of
intuitions. Initial intuitions arising in the context of prob-
lem-solving tend to be readily accessible conventional be-
liefs (Morewedge & Kahneman, 2010) that are associated
with a metacognitive feeling of rightness (Thompson,
Prowse Turner, & Pennycook, 2011) and appear to require
few cognitive resources (De Neys, 2006). Given the forego-
ing properties, initial intuitions during problem solving of-
ten pre-empt further analysis (Evans, 2008). Researchers
have constructed a number of tasks that present problem-
atic scenarios in which putatively objective information
conflicts with highly salient intuitions. Important for pres-
ent purposes, there appear to be substantial individual dif-
ferences in cognitive style (sometimes referred to as
thinking disposition), that is, the tendency to critically
evaluate initial misleading intuitions and persist in ana-
lytic processing (Stanovich, 2004). Focusing on individual
differences, we integrate theories of reasoning and deci-
sion-making with cognitive theories of religiosity and the
formation of belief and unbelief. We then test the hypoth-
esis that individual differences in religiosity can be pre-
dicted by individual differences in the propensity and
ability to question intuitions while solving reasoning
problems.
1.2. Cognitive styles and religiosity
The relation between analytic rationality and the rejec-
tion of religious beliefs has, of course, not gone unnoticed.
Atheists have generally been found to be, both stereotypi-
cally and empirically; intellectual, rational, and sceptical
(Beit-Hallahmi, 2006; Caldwell-Harris, Wilson, LoTempio,
& Beit-Hallahmi, 2010; Hunsberger & Brown, 2001). Intel-
lectualism has been found to be an important predictor of
religious apostasy among college students (Caplovitz &
Sherrow, 1977). Consistent with these attributes, atheists
most frequently give intellectual, rational, and scientific
reasons for their rejection of religious beliefs (Hunsberger
& Altemeyer, 2006). It is therefore hypothesized that when
intuitions conflict with reasoning, less religious people will
display a more analytic cognitive style than more religious
people. Moreover, in light of findings that skill in logical
inference is an important component of intelligence test-
ing (Stanovich & West, 2008), and the numerous studies
reporting a negative correlation between intelligence and
religiosity (e.g., Bertsch & Pesta, 2009; Larson & Witham,
1998; Lewis, Ritchie, & Bates, 2011; Lynn, Harvey, &
Nyborg, 2009; Reeve, 2009), it appears that highly religious
people may also be less skilled at basic logical inference
than less religious people. Therefore, based on the forgoing
evidence and reasoning, we hypothesize that more reli-
gious people, compared to less religious people, may be
both less skilled at logical inference (cognitive ability) as
well as more prone to be misled by immediate intuitions
(cognitive style) that essentially foreclose on the logical
processes that might draw inferences that would weaken
them.
Shenhav, Rand, and Greene (2011), working from a
dual-process framework, recently reported evidence con-
sistent with this hypothesis. In a series of studies run inde-
pendently of the current work, Shenhav and colleagues
demonstrated that performance on a reasoning task associ-
ated with analytic processing (i.e., the cognitive reflection
test or CRT; Frederick, 2005) was negatively correlated
with belief in God. CRT problems are structured to suggest
obvious but misleading answers to otherwise elementary
arithmetic questions and hence require further processing
to reject the incorrect answer that comes immediately to
mind. Shenhav and colleagues report that subjects failing
to reject the incorrect answer were more likely to believe
in God. Crucially, the correlation remained significant even
when cognitive ability (as measured by the Shipley Vocab-
ulary Test and the Wechsler Adult Intelligence Scale Matrix
Reasoning test) was controlled. Shenhav and colleagues
theorized that belief in God is predicted by reasoning style
because it is a particularly fundamental intuitive belief
(Atran, 2002; Barrett, 2004; Boyer, 1994; Guthrie, 1993)
and therefore hard to override via analytic processing.
There are, however, problems with this argument. First,
the authors cited by Shenhav and colleagues also fre-
quently describe religious beliefs as attractive and memo-
rable because they are ‘‘minimally’’ counterintuitive
(Barrett, 2000; Boyer, 2001; Norenzayan, Atran, Faulkner,
& Schaller, 2006; Pyysiäinen & Anttonen, 2002). Moreover,
as noted above, research on nonbelievers reveals that
increasing numbers of people find many if not all religious
concepts strongly counterintuitive (Hunsberger & Alte-
meyer, 2006). Finally, the interfering intuitions of the cog-
nitive tasks employed in dual process research, including
the CRT task used by Shenhav and colleagues, are not fun-
damental intuitions but merely plausible solutions to spe-
cific problems. Thus, while religious intuitions may or may
not be unique, we suggest a possibly more fundamental
basis for a negative association of analytic cognitive style
and religious beliefs: the asymmetry of belief and unbelief.
336 G. Pennycook et al. / Cognition 123 (2012) 335–346
1.3. The asymmetry of belief and unbelief
The asymmetric model of belief and unbelief posits that
comprehension automatically implies belief (Bain, 1859;
Gadamer, 1960;Spinoza, selected letters (S. Feldman Ed.;
S. Shirley, & IN: Hackett (Originally published in, 1677;
Gilbert, 1991; Gilbert, 1993). On this view, to understand
something is to implicitly accept it, at least briefly, as a pre-
requisite to understanding. It then requires a second move
to critically evaluate and certify or, alternatively, to ‘‘unbe-
lieve’’ it. Belief is therefore rapid, automatic, and effortless,
whereas the act of unbelief is slow, deliberate, and effort-
ful. Consistent with this, experimental manipulations
involving brief interruption of the opportunity to evaluate
the truth and falsity of beliefs leads consistently to a posi-
tive belief bias (Gilbert, Krull, & Mallone, 1990; Gilbert,
Tafarodi, & Malone, 1993). An asymmetry of belief and
unbelief hypothesis is clearly compatible with dual process
theories of problem solving and decision making. That is,
when a candidate answer to a problem spontaneously oc-
curs, the default, according to belief–unbelief asymmetry,
is initially to accept that answer as correct. It requires fur-
ther processing to cast doubt on the initial answer. Further
analysis is effortful, however, and must proceed in the face
of what seems, at least superficially, an already available
answer.
1.4. The present argument and hypotheses
A dual-process belief–unbelief asymmetry argument
relating analytic cognitive style to religious beliefs need
not require that religious beliefs are uniquely intuitive.
Rather, the proposal is that an analytic style of thinking
examines truth claims critically, though necessarily after
overcoming initial acceptance. All ideas are ultimately
open to question and analysis. Given belief–unbelief asym-
metry, however, unexamined ideas – or lightly examined
ideas – are unlikely to be doubted or rejected. Further anal-
ysis does, however, open the possibility of unbelieving,
modifying, or, of course, continued or even deeper accep-
tance of those ideas. Individuals with an analytic cognitive
style should therefore be more likely to overcome the
acceptance bias and reject or modify what they deem to
be unwarranted ideas, whether encountered in the course
of problem-solving, examining options in decision making,
or considering the truth-value of ideas more generally.
That religious beliefs are likely to be especially vulnerable
in a modernist context in which rationality and empiricism
provide the basis for truth claims is generally accepted by
both theists and atheists.
1
Thus, we suggest that individuals
with an analytic cognitive style are more likely to reject reli-
gious beliefs simply because such beliefs are vulnerable to
analytic evaluation.
The present argument applies not only to the basic be-
lief in God but also to any religious belief, such as a belief in
miracles or an afterlife. Moreover, an analytic stance, as
noted, need not lead to outright rejection but also to doubt
or to the modification of ideas. An analytic stance should
also lead to less conventional God beliefs (in Western soci-
ety), which might be thought of as ranging from orthodox
beliefs in a personal God, to notions of an abstract spirit or
force of nature, through doubt in God, and finally to out-
right rejection.
In the studies reported below we assess cognitive style
(analytic versus intuitive) using two reasoning tasks that
engender intuitive, but erroneous, responses. We assess
cognitive ability using a verbal intelligence measure as
well as a parallel reasoning task that does not imply a con-
flicting intuitive response. We also assess religious engage-
ment (R
e
) along with specific conventional religious beliefs
(R
b
) under the assumption that people who are engaged –
that is, they go to church, pray, and report that religion is
important in their everyday lives – are likely to do so, in
part, because of their religious beliefs. We postulate that
analytic reasoning acts directly on specific religious beliefs,
and that weakening of religious beliefs should then lead to
reduced religious engagement, including participation in
religious activities and rituals; that is, an analytic cognitive
style ?diminished religious belief ?reduced religious partic-
ipation relation. We therefore test a mediation hypothesis
in which an analytic thinking style directly predicts re-
duced religious belief, which then predicts reduced reli-
gious engagement fully explaining the expected bivariate
correlation between analytic cognitive style and religious
engagement. We also examine separately the association
of an analytic cognitive style with the nature of God beliefs.
In Study 2, we also assess paranormal forms of super-
natural belief. To the extent that paranormal beliefs are,
like religious beliefs, vulnerable to rational and empirical
disconfirmation, their rejection should be related to a more
analytic cognitive style. Preliminary evidence for this claim
comes from self-report data (Aarnio & Lindeman, 2005;
Aarnio & Lindeman, 2007; Lindeman & Aarnio, 2006; Lind-
eman & Aarnio, 2007) using the Rational–Experiential
Inventory (Pacini & Epstein, 1999). This research indicates
that paranormal believers report having more intuitive and
less analytic personality types (although without control-
ling for cognitive ability). Paranormal beliefs are less pre-
valent than religious beliefs and likely much less
generally intuitively appealing than religious beliefs (Rice,
2003). If the intuitiveness of religious belief is the crucial
factor that determines the relation between reasoning
style and religious belief, as suggested by Shenhav et al.
(2011), one would expect a weaker correlation between
reasoning style and paranormal belief. We therefore assess
both paranormal and religious belief and their relations to
performance measures of analytic reasoning while control-
ling for cognitive ability.
2. Study 1: religious belief
2.1. Method
2.1.1. Participants
Two hundred thirty-seven participants were recruited
through Mechanical Turk™. Two participants were ex-
1
Indeed, sophisticated theological arguments seldom attempt to defend
religious beliefs within an enlightenment scientific naturalism, but rather
prefer to critique the perceived narrowness of truth-tests via rationality,
efficient causation, and empiricism (e.g., Chapp, 2011; Gadamer, 1960).
G. Pennycook et al. / Cognition 123 (2012) 335–346 337
cluded because they failed an attention check question
presented half way through the procedure. For this, partic-
ipants were shown a list of activities and asked to write ‘‘I
read the instructions’’ in the ‘‘other’’ box if they were, in
fact, reading the instructions. Twelve more were dropped
from the sample because of missing information leaving
223 with complete data (131 females; average
age = 34.65, SD = 12.55), except for education which was
assessed for only 181 participants.
2
Participation was vol-
untary and participants received remuneration. All partici-
pants indicated that they lived in the United States.
Sessions lasted approximately 15 min.
2.1.2. Procedure
Participants first provided demographic information
(sex, age, education, and location). Cognitive variables
were assessed before asking questions concerning religious
belief to ensure that cognitive measures were not influ-
enced by participants’ hypotheses concerning the relation
between religion and decision making.
2.1.3. Cognitive measures of analytic cognitive style and
cognitive ability
Two different measures of analytic cognitive style (ACS)
were employed to provide some generality to our conclu-
sions. We, like Shenhav et al. (2011), used the cognitive
reflection test (CRT; Frederick, 2005); a widely used mea-
sure of cognitive ability and cognitive style (e.g., Toplak,
West, & Stanovich, 2011). The CRT consists of three
quasi-mathematical problems that generate implicit
misleading intuitions, for example:
A bat and a ball cost $1.10 in total. The bat costs $1.00
more than the ball. How much does the ball cost?
We also used three Base-Rate Conflict (BRC) problems
that contain a conflict between a salient stereotype and
more analytical probabilistic information. Consider the fol-
lowing BRC problem (taken from De Neys & Glumicic, 2008):
In a study 1000 people were tested. Among the
participants there were 995 nurses and 5 doctors. Jake is
a randomly chosen participant of this study.
Jake is 34 years old. He lives in a beautiful home in a
posh suburb. He is well spoken and very interested in pol-
itics. He invests a lot of time in his career.What is most
likely?
(a) Jake is a nurse.
(b) Jake is a doctor.
Here there are two conflicting pieces of information:
the base-rate probability of group membership (i.e.,
99.5% chance that Jake is a nurse) and diagnostic informa-
tion that cues an intuitive response based on stereotypical
beliefs about doctors and nurses (i.e., that Jake is a doctor).
Many people select the intuitive response and neglect or
underweight the base-rate information (De Neys &
Glumicic, 2008; for a review, see Barbey & Sloman,
2007). Those who are more willing (cognitive style) and
better able to (cognitive ability) engage in analytic process-
ing are less likely to select the intuitive response (Stano-
vich, 2009; Stanovich & West, 2000). Note that, unlike
the CRT, there is no objectively correct response for these
base-rate problems. Performance refers to the proportion
of non-intuitive (base-rate-consistent) responses selected.
As the ACS measures reflect both cognitive ability and
cognitive style, two measures were used to control for cog-
nitive ability (CA) independent of interfering intuitions.
WordSum is a brief vocabulary test that correlates well
with full scale measures of intelligence (Huang & Hauser,
1998). For this, participants were presented a list of 10 tar-
get words in capital letters and asked to choose the option
that most closely matches the meaning of the target word.
The second CA measure consisted of three Base-Rate Neu-
tral (BRN) problems that were structurally identical to the
BRC problems except that the personality descriptions
were not stereotypically associated with either response
alternative. Thus, BRN problems assess the basic cognitive
ability required to use the base-rate information to solve
the problem without the interfering intuitive response.
BRC problems, on the other hand, were intended to assess
the ability to override a conflicting and, in the context of
the problem, misleading, intuitive response. Thus, perfor-
mance on BRC problems while controlling for BRN perfor-
mance provides a measure of preference for probability
information over the stereotype independent of the ability
to use probability information, and is therefore reflective of
an analytic cognitive style. All cognitive variables were
converted to POMP scores (Cohen, Cohen, Aiken, & West,
1999) and unweighted means of the ACS and CA variables
were computed separately.
2.1.4. Measures of religiosity: religious engagement, religious
belief, and theistic belief
The religious engagement (R
e
) scale is intended to index
religious engagement or level of participation and consists
of three questions: importance of religion, church atten-
dance, and prayer frequency. The Religious Belief (R
b
) scale
assessed six conventional religious beliefs widely held by
religious people (International Social Survey Program
(ISSP), 1991, 1993). These are beliefs in heaven, hell, mira-
cles, afterlife, angels and demons, and an immaterial soul.
Each of the religiosity items was converted to a POMP
score, and a mean was then calculated across items from
each scale. Higher scores reflect higher R
e
and R
b
. A confir-
matory factor analysis supported separate R
e
and R
b
factors
(see Figs. S1 and S2 in the Supplementary materials). Both
scales had good internal consistency: Cronbach’s
a
= 0.87
for R
e
,
a
= 0.93 for R
b
. All items are listed in the Supplemen-
tary materials (Religiosity Inventory).
A third variable, a qualitative theistic belief measure
was based on the answer to the question: ‘‘What sort of
God, if any, do you believe in?’’ for which six options were
provided (proportion selected for each in brackets): (1)
Personal God (40.5%), (2) God as an impersonal force
(8.4%), (3) a God who created everything, but does not
intervene in human affairs (16.5%), (4) do not know
whether or not any of these Gods exist (12.7%), (5) do
2
OLS regression analyses reported in Tables 2, 4 and 5 were conducted
using list-wise deletion. The final models in these tables were also assessed
using imputation of missing data with full information maximum likeli-
hood structural equation modelling with effectively identical results (see
Fig. S3 in the Supplementary materials).
338 G. Pennycook et al. / Cognition 123 (2012) 335–346
not know whether or not any of these Gods exist, and no
one else does either (6.8%), and (6) I do not believe in Gods
of any sort (14.3%).
2.1.5. Demographic variables
Sex, (Males = 1, Females = 2), age in years (range: 18–
88), range of conservative versus liberal political ideology
[A five-point Likert scale with conservatism as the positive
pole: 1 = strongly liberal (16.9%), 2 = liberal (36.3%),
3 = moderate (21.9%), 4 = conservative (20.7%), 5 = strongly
conservative (4.2%)], and level of education [1 = some high
school or less(1.3%), 2 = high school (9.3%), 3 = technical,
trade or vocational training (3.0%), 4 = some college
(28.3%), 5 = college degree (27.0%), 6 = Master’s degree
(5.9%), 7 = Doctoral degree (1.3%)] were also assessed and
used as control variables. The present sample was rela-
tively highly educated and more liberal than the current
general US population (e.g., Gallup, 2011).
2.2. Results and discussion
Correlations among religious and cognitive variables,
along with sex, age, political ideology, and education are
presented in Table 1. Correlations of all cognitive variables
with R
b
were negative and significant. Sex and conserva-
tive political ideology were significantly positively corre-
lated with both religious measures and two of the
cognitive measures. Age was significantly correlated only
with WordSum. Education was significantly positively cor-
related with cognitive variables and age, and negatively
correlated with conservative political ideology.
As the theistic beliefs were categorical and, given the
likelihood of many ties with a small number of categories,
large sample gamma coefficients were used to assess the
association between theistic beliefs and cognitive variables
(
c
=0.22,0.13., 0.30, and 0.15, for WS, BRN, CRT, and
BRC, respectively, significant correlations in bold, p< .05).
To test the hypothesis of an independent effect of ana-
lytic cognitive style (ACS) on religious beliefs (R
b
), hierar-
chical multiple regression analyses were next carried out
predicting R
b
from ACS (unweighted mean of POMP scores
for CRT and BRC) while controlling for cognitive ability
(CA; unweighted mean of POMP scores for WS and BRN),
sex, age, conservative political ideology, and education
(Table 2). All demographic variables were entered first, fol-
lowed by R
e.
Finally, the cognitive measures were entered
simultaneously. Sex and conservatism were significant
independent predictors in step 1. In step 2, R
e
made a sig-
nificant independent contribution to the prediction of R
b
while sex and conservatism remained significant. Interest-
ingly, the beta increased for education, becoming margin-
ally significant, suggesting that more highly educated
individuals may be less likely to accept specific religious
beliefs than less educated individuals at equivalent levels
of religious engagement. In step three, ACS, but not CA,
made a further significant independent contribution, and
both sex and conservatism remained significant. In a sepa-
rate analysis, entering CA separately from ACS, the beta for
CA was non-significant. Thus, higher levels of analytic cog-
nitive style predicted religious beliefs independently of
sex, political ideology, education, R
e
, and cognitive ability.
At each step, R
2
change
indicated a significant improvement
in the prediction of R
b
.
The results of Study 1 provide evidence that an analytic
tendency to ignore or override initial intuitive responses
engendered by either implicit assumptions (CRT) or expli-
cit stereotypes (BRC) is a reliable predictor of religious be-
lief independent of cognitive ability or other control
variables. This extends Shenhav et al.’s (2011) finding that
reasoning style predicts one particular religious belief (be-
lief in God) to a broader set of religious beliefs (heaven,
hell, miracles, afterlife, angels and demons, and an imma-
terial soul). Moreover, the results for a sample of qualita-
tively different theistic beliefs suggest a more nuanced
association than mere degree of belief. That is, analytic
cognitive style predicted different degrees or graded kinds
of God belief, from personal God to atheism (discussed in
detail below).
In Study 2, we attempt to replicate the effects found in
Study 2 with a different population. We also test the
hypothesis that an analytic cognitive style also predicts
diminished acceptance of paranormal beliefs. More gener-
ally, based on the asymmetric model of belief and unbe-
lief hypothesis, we expected measures of paranormal
beliefs to behave similarly to the specific religious beliefs
variable.
Table 1
Pearson product–moment correlations among major variables in Study 1. CA – cognitive ability, ACS – analytic cognitive style, R
e
– religious engagement, R
b
religious beliefs, WS – WordSum, BRN – base-rate neutral, CRT – cognitive reflection test, BRC – Base-rate conflict, sex (Male = 1, Female = 2), age, conservatism,
education. N= 231, except for education for which N= 181. Coefficients in bold are significant, p< .05.
R
b
CA ACS Demographic variables
WS BRN CRT BRC Sex Age Conservatism Education
R
e
0.77 0.26 0.09 0.23 0.07 0.18 0.02 0.38 0.07
R
b
0.24 0.15 0.33 0.19 0.27 0.00 0.46 0.05
WS 0.26 0.37 0.14 0.14 0.28 0.14 0.26
BRN 0.26 0.32 0.08 0.04 0.08 0.18
CRT 0.26 0.16 0.12 0.16 0.19
BRC 0.06 0.05 0.12 0.15
Sex 0.12 0.14 0.02
Age 0.03 0.14
Conservatism 0.13
G. Pennycook et al. / Cognition 123 (2012) 335–346 339
3. Study 2: religious and paranormal beliefs
3.1. Method
3.1.1. Participants
Two hundred and eighty-seven participants were re-
cruited through email and completed the study online.
Twenty participants were excluded because they submit-
ted incomplete data or failed an attention check question
that occurred half way through the procedure. The partic-
ipants (208 females; average age = 35.04, SD = 12.77) had
left their email address after completing a survey as part
of an ongoing project concerning sleep paralysis (Cheyne,
2010). This sample, in contrast to the US sample in Study
1, was broadly international, including participants from
USA (38.7%), Canada (24.7%), United Kingdom (24%), Ocea-
nia, (9.3%), Europe, (2.7%) and various other regions (0.7%).
All participants indicated that their primary language was
English. Sessions lasted approximately 40 min and partici-
pation was voluntary.
3.1.2. Procedure and materials
Religious and cognitive variables were identical to Study
1. As in Study 1, participants completed the cognitive mea-
sures before being asked about their religious beliefs. Par-
ticipants were also given a slightly revised version of the
Paranormal Belief Scale (Tobacyk, 1988; original scale:
Tobacyk & Milford, 1983), in that we excluded the redun-
dant religious items from the religious belief subscale of
the Paranormal Belief Scale. We used the remaining six sub-
scales (example items in parentheses): Psi (‘‘Mind reading
is possible’’), Witchcraft (‘‘Witches do exist’’), Omens of
luck (‘‘Black cats can bring bad luck’’), Spiritualism (‘‘It is
possible to communicate with the dead’’), Extraordinary
life forms (‘‘The Loch Ness monster of Scotland exists’’)
and Precognition (‘‘Astrology is a way to accurately predict
the future’’). Scores on the subscales were summed to cre-
ate an overall paranormal belief (P
b
) score. All three scales
had acceptable internal consistency, Cronbach’s
a
= 0.84,
0.89, and 0.96, for R
e
,R
b
, and P
b
, respectively.
Demographic variables were assessed and coded as in
Study 1. The age range was from 16 to 69. Conservative
political ideology scores again indicated a relatively liberal
sample overall: strongly liberal (28.7%), liberal (36.2%),
moderates (21.3%), conservative (9.9%), strongly conserva-
tive (3.9%). The sample was also, again, relatively highly
educated. Most of the sample indicated having a Bachelor’s
degree (30%), a Master’s degree (11.3%), a Doctoral degree
(2%), or some college experience but no degree (30%). The
remainder had vocational training (10.7%), high school
(8%), or some high school or less (4.1%).
3.2. Results and discussion
Correlations among the major variables are presented
in Table 3.R
e
,R
b
, and P
b
were significantly positively corre-
lated. All cognitive variables were significantly correlated
with R
b
and all cognitive variables except WS were signif-
icantly correlated with R
e
and P
b
. Sex was again correlated
with both religious and cognitive variables though some-
what less consistently so than in Study 1. Conservative
political ideology was again consistently significantly cor-
related with religious variables and also with P
b
, and con-
sistently but weakly and generally non-significantly with
cognitive variables. Education was significantly negatively
correlated with P
b
and conservative political ideology and
positively with cognitive variables.
Overall, the results for R
b
and P
b
were, as expected,
highly similar. We note that, although P
b
and R
b
were sig-
nificantly correlated, only 14.7% of the sample could be
considered consistent believers in the supernatural (i.e.,
in the top 3rd on both P
b
and R
b
scales). Indeed, P
b
and
Table 2
Hierarchical multiple regression analyses predicting religious beliefs (R
b
) with analytic cognitive style (ACS) controlling for sex, age, conservatism, education,
religious engagement (R
e
) and cognitive ability (CA).
B S.E. btp
R
2
D
p
D
Intercept 33.42 8.98 3.72 0.000
Sex 12.15 3.00 0.28 4.05 0.000
Age 0.08 0.12 0.05 0.67 0.501
Conservatism 7.22 1.27 0.39 5.67 0.000
Education 0.36 1.25 0.02 0.29 0.774
0.26 0.001
Intercept 36.74 6.35 5.79 0.000
Sex 7.42 2.15 0.17 3.45 0.001
Age 0.02 0.09 0.01 0.26 0.796
Conservatism 2.87 0.96 0.15 2.99 0.003
Education 1.65 0.89 0.09 1.86 0.065
R
e
0.46 0.04 0.67 12.91 0.000
0.38 0.001
Intercept 40.02 6.99 5.73 0.000
Sex 6.49 2.20 0.15 2.95 0.004
Age 0.06 0.09 0.03 0.66 0.511
Conservatism 2.77 0.94 0.15 2.95 0.004
Education 0.77 0.93 0.04 0.83 0.407
R
e
0.44 0.04 0.64 12.35 0.000
CA 0.04 0.07 0.03 0.52 0.602
ACS 0.11 0.04 0.14 2.52 0.013
0.02 0.013
340 G. Pennycook et al. / Cognition 123 (2012) 335–346
R
b
were correlated only for the participants in the lowest
3rd on the P
b
(r= .564, p< .001) and R
b
(r= .500, p< .001)
scales (all other p’s > .162). Indeed, the P
b
and R
b
correla-
tion was linear only over the first two-thirds of the R
b
scale
(see Fig. S3, Supplementary materials). Thus, consistent
with previous research (Aarnio & Lindeman, 2007), the
large majority of believers tend to be of the paranormal
or religious type, but not both. The relatively weak correla-
tion between cognitive ability and P
b
is consistent with
previous research indicating only a tenuous correlation be-
tween intelligence and paranormal belief (Smith, Foster, &
Stovin, 1998).
We again examined the association between cognitive
performance and theistic beliefs. The coefficients for CRT
and BRC are highly similar to those in Study 1,
c
=.06,
.22,.26, and .18, for WS, BRN, CRT, and BRC, respec-
tively (significant correlations in bold, p< .05). To examine
the association of the qualitative differences among theis-
tic beliefs and analytic cognitive style we combined the
data from Studies 1 and 2 and cross-tabulated the six dif-
ferent theistic beliefs with CRT performance (i.e., whether
a participant answered 0, 1, 2, or 3 CRT questions cor-
rectly). This created a 4 6 table with cell frequencies
ranging from 7 to 96 with only two cells with less than
10 observations. Analysis of this table yielded a significant
likelihood-ratio
v
2
(15) = 42.71, p< .001, and a gamma
coefficient comparable to those computed for the two
studies separately,
c
= .26, p< .001. Fig. 1 illustrates the
negative association between theistic beliefs and CRT per-
formance. Examination of the adjusted standardized resid-
uals indicated that those expressing belief in a personal
God were significantly overrepresented, and atheists sig-
nificantly underrepresented among participants answering
none of the CRT questions correctly, whereas the pattern
was reversed among those answering two or three ques-
tion correctly. Over half of those reporting the most con-
ventional God belief (a personal God) scored 0 out of 3
on the CRT, whereas over 60% of atheists answered at least
2 of 3 questions correctly. The residual differences indi-
cated that significantly more believers in a personal God
and fewer atheists answered zero questions than partici-
pants in all other categories (all p’s < .001). Also, signifi-
cantly fewer universal agnostics obtained zero scores
than Deists (all p’s < .01). At the other end of the CRT scale,
atheists were significantly more likely to be among those
answering all three questions correctly than participants
from all other categories (all p’s < .005), and believers in a
personal God and Pantheists were less likely to do so (all
p’s < .05). Moreover, the rather orderly pattern of scores
suggests that these categories might represent a scale of
conventionality/unconventionality of belief from: (1) a cul-
turally conventional personal God, though (2) a Spinozist
Pantheism, (3) Deism, (4) personal and (5) universal agnos-
ticism, to (6) atheism. These results also suggest that an
analytic stance is associated with more than simple accep-
tance or rejection of belief in God.
Hierarchical regression analyses predicting R
b
with ana-
lytic cognitive style (ACS) controlling for sex, education, P
b
and cognitive ability (CA) replicated the main findings of
Study 1 (Table 4). Sex and conservatism significantly inde-
pendently predicted R
b
in step 1. R
e
was highly significant
and sex remained so at step 2. At step 3, ACS, but not CA,
made a significant contribution to the prediction of R
b
along with sex and R
e
.R
2
change
was highly significant at each
step. As in the previous regression analysis, we entered CA
separately from ACS and the beta for CA was non-
significant.
A regression analysis predicting paranormal beliefs par-
allel to that for R
b
(Table 5) was carried out next. At the
Table 3
Pearson product–moment correlations among major variables in Study 2. R
e
– religious engagement, R
b
– religious beliefs, P
b
– paranormal beliefs, WS –
WordSum, BRN – base-rate neutral, CRT – cognitive reflection test, BRC – base-rate conflict, sex, age, conservatism, education. N= 267. Coefficients in bold are
significant, p< .05.
R
b
P
b
CA ACS Demographic variables
WS BRN CRT BRC Sex Age CPI ED
R
e
0.74 0.32 0.05 0.27 0.23 0.20 0.17 0.10 0.29 0.00
R
b
0.55 0.13 0.22 0.29 0.31 0.25 0.04 0.28 0.09
P
b
0.06 0.13 0.31 0.23 0.22 0.05 0.02 0.10
WS 0.11 0.24 0.20 0.05 0.24 0.10 0.25
BRN 0.17 0.43 0.09 0.12 0.10 0.28
CRT 0.19 0.23 0.02 0.06 0.18
BRC 0.04 0.10 0.15 0.25
Sex 0.12 0.00 0.04
Age 0.13 0.00
CPI 0.19
Fig. 1. The number of correct responses on the three item cognitive
reflection test as a function of type of theistic belief. Asterisks indicate
cells with significant adjusted residuals, p< .02.
G. Pennycook et al. / Cognition 123 (2012) 335–346 341
first step, only sex made a significant independent contri-
bution to P
b
. At step 2, R
e
made a significant independent
contribution to the prediction of P
b
. Sex remained signifi-
cant and education became marginally significant step 2.
At step 3, ACS, but not CA, made a significant independent
contribution to the prediction of P
b
along with sex and R
e
.
R
2
change
was significant at each step. Again, we entered CA
separately from ACS and the beta for CA was non-signifi-
cant. Not surprisingly, the association between ACS and
P
b
, was less attenuated by R
e
(Table 4) than was the asso-
ciation between ACS and R
b
(Table 5).
In summary, the regression analyses presented in Ta-
bles 3–5 indicated that analytic cognitive style made con-
sistently significant contributions to both religious and
paranormal beliefs over and above cognitive ability and
the included demographic variables. However, given the
significant positive correlation between the two types of
supernatural belief, it is possible that the analyses for R
b
and P
b
are simply redundant analyses. To assess the inde-
pendence of the relation between ACS and each of the
supernatural belief variables (R
b
and P
b
) a separate regres-
sion analysis was conducted entering both belief variables
Table 4
Multiple regression analyses predicting religious beliefs (R
b
) with analytic cognitive style (ACS) controlling for sex, age, conservatism, education, religious
engagement (R
e
), paranormal beliefs (P
b
), and cognitive ability (CA).
B S.E. btp
R
2
D
p
D
Intercept 6.55 10.53 0.62 0.535
Sex 15.42 3.50 0.25 4.40 0.000
Age 0.08 0.11 0.04 0.71 0.476
Conservatism 6.20 1.36 0.27 4.57 0.000
Education 0.47 1.08 0.03 0.43 0.665
0.14 0.001
Intercept 20.51 7.45 2.75 0.006
Sex 7.47 2.51 0.12 2.98 0.003
Age 0.04 0.08 0.02 0.52 0.606
Conservatism 1.47 1.00 0.06 1.48 0.140
Education 1.34 0.76 0.07 1.76 0.080
R
e
0.66 0.04 0.71 16.42 0.000
Intercept 23.31 8.46 2.76 0.006 0.44 0.001
Sex 6.08 2.47 0.10 2.46 0.014
Age 0.07 0.08 0.03 0.85 0.397
Conservatism 1.46 0.97 0.06 1.50 0.134
Education 0.61 0.81 0.03 0.76 0.447
R
e
0.62 0.04 0.67 15.22 0.000
CA 0.05 0.08 0.03 0.60 0.550
ACS 0.16 0.04 0.18 3.91 0.000
0.02 0.001
Table 5
Multiple regression analyses predicting paranormal beliefs (P
b
) with analytic cognitive style (ACS) controlling for sex, age, conservatism, education, religious
engagement (R
e
), and specific religious beliefs (R
b
), and cognitive ability (CA).
B S.E. btp
R
2
D
p
D
Intercept 12.79 8.66 1.48 0.141
Sex 10.91 2.88 0.23 3.79 0.000
Age 0.13 0.09 0.08 1.33 0.186
Conservatism 0.23 1.12 0.01 0.20 0.840
Education 1.33 0.89 0.09 1.50 0.136
0.06 0.002
Intercept 17.65 8.33 2.12 0.035
Sex 8.15 2.81 0.17 2.90 0.004
Age 0.08 0.09 0.05 0.91 0.365
Conservatism 1.87 1.12 0.10 1.68 0.095
Education 1.64 0.85 0.11 1.92 0.056
R
e
0.23 0.04 0.31 5.11 0.000
0.09 0.001
Intercept 18.55 9.40 1.97 0.050
Sex 6.48 2.75 0.14 2.36 0.019
Age 0.05 0.09 0.03 0.54 0.587
Conservatism 1.89 1.08 0.10 1.75 0.081
Education 0.90 0.90 0.06 1.01 0.315
R
e
0.19 0.05 0.26 4.18 0.000
CA 0.10 0.08 0.08 1.17 0.243
ACS 0.19 0.04 0.28 4.38 0.000
0.06 0.001
342 G. Pennycook et al. / Cognition 123 (2012) 335–346
simultaneously as predictors of ACS. Betas for both belief
variables were significant, b(R
b
)=0.27, p< . 001,
b(P
b
)=.20, p< .002, R
2
= .17, F(2, 279) = 29.05, p< .001.
Thus, the results for each type of supernatural belief are
independent tests of the analytic cognitive style Msuper-
natural beliefs relation.
3.3. Testing the ACS ?R
b
?R
e
mediation hypothesis (Studies
1 and 2)
Structural equation modeling of combined data from
both studies was used to test the hypothesis that the asso-
ciation between analytic cognitive style and religious
engagement is mediated by conventional religious beliefs.
Study 1 and Study 2 data were entered as groups in a struc-
tural equation modeling analysis in which ACS predicted
R
b
and R
e
and R
b
predicted R
e
. To assess the replication of
mediation across both studies, the unstandardized coeffi-
cient for the path from ACS to R
e
was set to zero for both
groups. This constrained model was well-fitting by several
goodness-of-fit indices, including a non-significant chi-
square,
v
2
(2) = 3.31, p= .191; small root mean square er-
ror, RMSEA = .035; and robust Bentler–Bonnett normed
fit, NFI = .994 and comparative fit-indices, CFI = .997. Sepa-
rate mediation analysis of the data from each study yielded
highly similar standardized path coefficients (see Fig. 2).
There were also robust indirect effects for ACS on R
e
, for
both samples, 0.29 and 0.28, p’s < .001 for study 1 and
study 2 respectively. In both samples the mediation effect
was also significant by Sobel tests, z= 5.30 and 6.54,
respectively, p’s < .001.
4. General discussion
The present research provides evidence that an analytic
cognitive style, defined as a propensity to engage in effort-
ful reasoning, is associated with a tendency to subscribe to
both religious and paranormal forms of supernatural belief.
Participants likely to reject an intuitive response for two
types of reasoning problems were more likely to reject
supernatural beliefs, even when cognitive ability, as well
as sex, age, political ideology, education and religious
engagement were controlled.
The present studies, completed independently from
those reported in Shenhav et al. (2011), replicate and ex-
tend their reported relation between reasoning style and
theistic belief. Shenhav and colleagues based their predic-
tions on the assumption that religious beliefs are uniquely
intuitive and scored the CRT in terms of ‘‘intuitive’’ re-
sponses, which are highly, though not perfectly, correlated
with correct responses based on analytic reasoning; so
highly (r=0.90 in each of our two studies) as to be indis-
tinguishable in terms of predictive power. Our argument
and approach is, however, somewhat different. Even if it
is indeed the case that some individuals reason about
supernatural beliefs more analytically than others, why
does increased analytic processing so systematically de-
crease the likelihood of supernatural belief? In our view,
beliefs in things such as angels, demons, black magic, or
mind reading are unlikely to be reinforced by analytic
thought, not because they are especially intuitive, but be-
cause they are counterintuitive; that is, violations of a nat-
uralistic worldview (Atran & Norenzayan, 2004). We
suggest therefore that analytic individuals have decreased
levels of supernatural belief because they are more likely
to scrutinize ideas, detect such violations, and unbelieve
them.
The present results regarding theistic beliefs, though
broadly in agreement with those of Shenhav and col-
leagues, are also somewhat more detailed with regard to
God belief and unbelief. What is perhaps the most striking
finding of the present study goes beyond a demonstration
of a negative relation between an analytic cognitive style
and degree of belief in God. This is the finding, illustrated
in Fig. 1, that increasing performance on the CRT was asso-
ciated with a systematically changing conception of God. It
was not simply the case that more analytic individuals
were more likely to reject all forms of God belief, though
many were, but that they may also sometimes adopt less
conventional and more abstract God beliefs. God as an ab-
stract force or a non-intervening creator is sufficiently
unconventional to require at least some analytic reflection
upon the nature of God. They also represent positions that
Fig. 2. Mediation of the analytic cognitive style – religious engagement relation by religions beliefs. Standardized path coefficients for Study 1 above lines
and for Study 2 below lines. Values in parentheses indicate bivariate regression coefficients for each study.
G. Pennycook et al. / Cognition 123 (2012) 335–346 343
arguably resolve contradictions between theism and a nat-
uralistic worldview.
None of the foregoing conclusions were qualified by the
demographic variables included in the present study.
Ancillary findings involving the demographic variables
were generally in agreement with previous research. Con-
sistent with previous research, women obtained higher
scores on all measures of religiosity (Argyle & Beit-Hallahmi,
1975; Batson, Schoenrade, & Ventis, 1993; Benson,
Donahue, & Erickson, 1989) and paranormal beliefs (Stark
& Bainbridge, 1985). Consistent with Frederick (2005),
women also obtained lower scores on the CRT in both
studies. Frederick also presented evidence that the CRT sex
difference is independent of math ability and that women
gave more intuitive responses relative to men when looking
entirely at incorrect responses. More specifically, there are
two ways to be wrong on the CRT: (1) by giving the incorrect
intuitive response and (2) by giving an incorrect non-
intuitive response, and a larger proportion of the males’
incorrect responses were of the second type. The present
results thus open the possibility that part of the sex differ-
ence in religiosity might be explained by sex differences in
analytic cognitive style (see also Aarnio & Lindeman, 2005).
Also consistent with previous research (Bouchard,
2009; Grupp & Newman, 1973; Keysar, 2007; Kosmin,
2008; Zelan, 1968) more conservative participants were
significantly more religious. Conservative participants
were also somewhat less analytic, though the effects were
generally non-significant. The effects for education were
also consistent with previous research, in that the associa-
tion with religion and conservative political ideology was
marginally negative (Albrecht & Heaton, 1984; Argyle &
Beit-Hallahmi, 1975; McFarland, Wright, & Weakliem,
2011; Petersen, 1994). The correlation between education
and our cognitive variables was much stronger, however,
as more highly educated individuals consistently scored
higher on measures of cognitive ability and analytic think-
ing style.
The present research findings agree with those of re-
search reporting a correlation between paranormal beliefs
and self-reported measures of intuitive versus analytic
thinking style (Aarnio & Lindeman, 2005; Aarnio &
Lindeman, 2007; Lindeman & Aarnio, 2006; Lindeman &
Aarnio, 2007). Also consistent with research on intelligence
(Bertsch & Pesta, 2009; Lynn et al., 2009; Reeve, 2009), we
found several significant negative correlations between
cognitive measures thought to be generally reflective of
cognitive ability – as assessed by neutral base-rate prob-
lems and WordSum – and religiosity (both engagement
and beliefs) as well as paranormal beliefs. Measures of
cognitive ability, however, produced weaker and less
consistent correlations with religiosity and paranormal
beliefs than did the analytic cognitive style measures.
Consistent with the mediation hypothesis (i.e., that cog-
nitive style differences predict religious behavior indirectly
via belief differences), not only did analytic cognitive style
predict religious and paranormal beliefs independently of
demographic variables, it did so independently of religious
engagement and, moreover, the effects on religious beliefs
completely explained the negative association between
analytic cognitive style and religious engagement. This is
critical for a theory of cognitive influences on religiosity;
namely that analytic cognitive style negatively affects reli-
giosity and does so via lower acceptance of conventional
religious beliefs. This is consistent with the Shenhav et al.
(2011) finding that CRT performance predicted self-re-
ported change in religious beliefs. Neither finding clearly
establishes causal direction, of course. Indeed, an intrigu-
ing speculative alternative explanation is that examining
one’s beliefs and finding them wanting is a motivator of
analytic thinking. It is worth noting, however, that Shen-
hav and colleagues also reported experimental evidence
for a short-term causal link between cognitive style and
belief in God by demonstrating diminished God belief for
participants who were primed to think analytically (and
vice versa for those who were primed to think intuitively).
Finding a short-term situational effect is intriguing in that
it suggests that religious beliefs are, to some extent at least,
reevaluated each time they are considered; a process con-
sistent with an asymmetrical belief–unbelief hypothesis.
The application of the belief–unbelief asymmetry mod-
el to religious and paranormal beliefs extends previous
work that focused on narratives about mundane facts
and secular events (Gilbert et al., 1990; Gilbert et al.,
1993) to religious and paranormal beliefs. The results are
consistent with (though not direct evidence for) the
hypothesis that supernatural belief is a default state and
that a consequence of analytic reasoning styles is to under-
mine default beliefs. Thus, on this view, people are biased
to believe in God, heaven, hell, angels, demons, souls, and
other supernatural beliefs embedded in the culture when
they first learn about those concepts but those disposed
to analyze such concepts will be more likely to change
their mind. Of course, further research is required to inte-
grate supernatural belief with the belief–unbelief asymme-
try model under an experimental design.
Given the hypothesis that supernatural belief is a de-
fault state that requires some level of analytic processing
to override and given the generally accepted proposition
that analytic reasoning is, on the whole, underutilized
(Evans, 2008; Kahneman, 2003; Stanovich, 2009), the high
levels of supernatural belief worldwide are unsurprising.
Perhaps what, on this view, needs explaining is decreasing
religiosity in many secular Western societies (e.g., Crockett
& Voas, 2006; Kosmin & Keysar, 2009; Zuckerman, 2008). It
is important to acknowledge that numerous social and cul-
tural factors will condition the vulnerability of supernatu-
ral claims to analytic reasoning, including institutional and
formal integration of specific supernatural beliefs into
coherent master narratives accepted by the larger culture.
Of course, the extent to which secularism and religious
pluralism continue to increase and cause traditional reli-
gious master narratives to lose their hegemony, the inte-
grative power and authority of such narratives will
decrease, causing them to become more vulnerable to
unbelief. We speculate, therefore, that the vulnerability
of beliefs to being unbelieved via analytic processing will
depend upon both the availability of contradictory beliefs
and the integration of such contradictory beliefs into an
authoritative formal system.
Future studies might therefore consider the different
degrees of acceptance of alternative cultural master narra-
344 G. Pennycook et al. / Cognition 123 (2012) 335–346
tives (religious versus secular-scientific) by individuals dif-
fering in analytic cognitive style. There are two interesting
possibilities. More highly analytic individuals may simply
be more sceptical of all epistemic claims independently
of worldview. Alternatively, more highly analytic individu-
als may not only reject religious perspectives but may also
be more attracted to a secular scientific worldview, given
that such a worldview claims to invite critical analysis as
an integral part of its ‘‘self-correcting’’ nature. In addition,
future studies might also investigate the role of scientific
beliefs themselves, which, ex hypothesi, would initially be
believed for the same (rapid, automatic, and effortless pro-
cessing) reasons as supernatural beliefs. The present model
makes no predictions, for example, about whether highly
analytic individuals will be more or less likely to initially
believe new scientific factual claims, but predicts that an
analytic thinking style should render individuals more
likely to abandon disconfirmed scientific beliefs or new sci-
entific facts at odds with prior facts and theory depending
upon their prior probabilities.
Considered in light of research suggesting an important
role for analytic thinking styles, the asymmetric belief–
unbelief hypothesis may be quite general. If beliefs must
first be accepted in order to be understood, as the available
data suggests (Gilbert, 1991, 1993), the propensity to hold
any set of empirically or logically vulnerable beliefs might
potentially be predicted by the ability and willingness to
think analytically. The present results clearly supported
the hypothesis that individual differences in analytic cog-
nitive style are useful for understanding selected cognitive
aspects of supernatural belief in particular. Our data are
consistent with the simple but, in our opinion, profound
idea that two people who share the same cognitive ability,
education, political ideology, sex, age and level of religious
engagement may acquire very different sets of beliefs
about the world if they differ in their propensity to think
analytically.
Acknowledgement
Funding for this study was provided by the Natural Sci-
ences and Engineering Research Council of Canada.
Appendix A. Supplementary material
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.cognition.2012.03.003.
References
Aarnio, K., & Lindeman, M. (2005). Paranormal beliefs, education, and
thinking styles. Personality and Individual Differences, 39, 1227–1236.
Aarnio, K., & Lindeman, M. (2007). Religious people and paranormal
believers: Alike or Different. Journal of Individual Differences, 28, 1–9.
Albrecht, S. L., & Heaton, T. B. (1984). Secularization, higher education,
and religiosity. Review of Religious Research, 26, 43–58.
Argyle, M., & Beit-Hallahmi, B. (1975). The social psychology of religion.
London, UK: Routledge and Kegan Paul.
Atran, S. (2002). In gods we trust: The evolutionary landscape of religion.
Oxford: Oxford University press.
Atran, S., & Norenzayan, A. (2004). Religion’s evolutionary landscape:
Counterintuition, commitment, compassion, communion. Behavioural
and Brain Sciences, 27, 713–770.
Bain, A. (1859). The emotions and the will. London: Longmans, Green.
Barbey, A. K., & Sloman, S. A. (2007). Base-rate respect: From ecological
rationality to dual processes. Behavioural and Brain Sciences, 30,
241–256.
Barrett, J. L. (2000). Exploring the natural foundations of religion. Trends in
Cognitive Science, 4, 29–34.
Barrett, J. L. (2004). Why would anyone believe in God? Lanham, MD:
Altamira.
Batson, C. D., Schoenrade, P., & Ventis, W. L. (1993). Religion and the
individual: A social-psychological perspective. New York, NY, US: Oxford
University Press.
Beit-Hallahmi, B. (2006). Atheists: A psychological profile. In M. Martin
(Ed.), The Cambridge companion to atheism. New York: Cambridge
University Press.
Benson, P. L., Donahue, M. J., & Erickson, J. A. (1989). Adolescence and
religion: A review of the literature from 1970–1986. Research in the
Social Scientific Study of Religion, 1, 153–181.
Bertsch, S., & Pesta, B. J. (2009). The Wonderlic Personnel Test and
elementary cognitive tasks as predictors of religious sectarianism,
scriptural acceptance, and religious questioning. Intelligence, 37,
231–237.
Bouchard, T. J. Jr., (2009). Authoritarianism, religiousness, and
conservatism: I ‘‘obedience to authority’’ the explanation for their
clustering, universality and evolution? In E. Voland & W.
Schiefenhöven (Eds.), The biological evolution of religious mind and
behaviour (pp. 165–180). New York, NY: Springer.
Boyer, P. (1994). The naturalness of religious ideas: A cognitive theory of
religion. Berkeley, CA: University of California Press.
Boyer, P. (2001). Religion explained: The evolutionary origins of religious
thought. New York, NY: Basic Books.
Caldwell-Harris, C. L., Wilson, A. L., LoTempio, E., & Beit-Hallahmi, B.
(2010). Exploring the atheist personality: Well-being, awe, and
magical thinking in atheists, Buddhists, and Christians. Mental
Health, Religion & Culture, 7, 659–672.
Caplovitz, D., & Sherrow, F. (1977). The religious drop-outs: Apostasy among
college students. Beverly Hills, CA: Sage Press.
Chapp, L. S. (2011). The God of covenant and creation: Scientific naturalism and
its challenge to the Christian faith. London, UK: T&T Clark International.
Cheyne, J. A. (2010). Recurrent isolated sleep paralysis. In M. Thorpy & G.
Plazzi (Eds.), Parasomnias and other movement-related sleep disorders.
Amsterdam: Elsevier.
Cohen, P., Cohen, J., Aiken, L. S., & West, S. G. (1999). The problem of units
and the circumstance for POMP. Multivariate Behavioral Research, 34,
315–346.
Crockett, A., & Voas, D. (2006). Generations of decline: Religious change in
20th Century Britain. Journal for the Scientific Study of Religion, 45,
567–584.
De Neys, W. (2006). Dual processing in reasoning – Two systems but one
reasoner. Psychological Science, 17, 428–433.
De Neys, W., & Glumicic, T. (2008). Conflict monitoring in dual process
theories of thinking. Cognition, 106, 1284–1299.
Epstein, S. (1994). Integration of the cognitive and psychodynamic
unconscious. American Psychologist, 49, 709–724.
Evans, J. B. T. (2008). Dual-processing accounts of reasoning, judgment,
and social, cognition. Annual Review of Psychology, 59, 255–278.
Frederick, S. (2005). Cognitive reflection and decision making. The Journal
of Economic Perspectives, 19, 25–42.
Frey, U. (2009). Cognitive Foundations of Religiosity. In: E., Voland & W.
Schiefenhövel (Eds.), The biological evolution of religious mind and
behavior (pp. 229–241). Heidelberg, DE: Springer.
Gadamer, G. (1960). Truth and method (second, revised edition). (J.
Weinsheimer & D. G. Marshall Trans.), New York, NY: Crossroad.
Gallup (2011). US Political ideology stable with conservatives leading.
August 1, <http://www.gallup.com/poll/148745/political-ideology-
stable-conservatives-leading.aspx>.
Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46,
107–119.
Gilbert, D. T., Krull, D. S., & Mallone, P. S. (1990). Believing the
unbelievable: Some problems in the rejection of false information.
Journal of Personality and Social Psychology, 59, 601–613.
Gilbert, D. T., Tafarodi, R. W., & Malone, P. S. (1993). You can’t not believe
everything you read. Journal of Personality and Social Psychology, 65,
221–233.
Gilbert, D. T. (1993). The assent of man: The mental representation and
control of belief. In D. M. Wegner & J. W. Pennebaker (Eds.), Handbook
of mental control (pp. 57–87). Englewood Cliffs, NJ: Prentice Hall.
G. Pennycook et al. / Cognition 123 (2012) 335–346 345
Grupp, F. W., Jr., & Newman, W. M. (1973). Political ideology and religious
preference: The John Birch Society and Americans for democratic
action. Journal for the Scientific Study of Religion, 12, 401–413.
Guthrie, S. E. (1993). Faces in the clouds: A new theory of religion. Oxford:
Oxford University Press.
Huang, M. H., & Hauser, R. M. (1998). Trends in Black–White test score
differentials: II the WORDSUM vocabulary test. In U. Neisser (Ed.), The
rising curve: Long-term gains in IQ and related measures (pp. 303–332).
Washington, DC: American Psychological Association.
Hunsberger, B. E., & Altemeyer, B. (2006). Atheists: A groundbreaking study
of America’s nonbelievers. Amherst, NY: Prometheus Books.
Hunsberger, B. E., & Brown, L. B. (2001). Religious socialization, apostasy
and the impact of family background. Journal for the Scientific Study of
Religion, 23, 239–251.
International Social Survey Program (ISSP), 1991, 1993, <http://
www.Chass.utoronto.ca/datalib/major/issp.htm>.
Kahneman, D. (2003). A perspective on judgment and choice: Mapping
bounded rationality. American Psychologist, 58, 697–720.
Keysar, A. (2007). Who are America’s atheists and agnostics? In B. Kosmin
& A. Keysar (Eds.), Secularism and secularity: Contemporary
international perspectives (pp. 33–39). Hartford, CT: Institute for the
Study of Secularism in Society and Culture.
Kosmin, B. (2008). A religious, irreligious and anti-religious Americans:
The no religion population of the US – Nones. <http://www.
trincoll.edu/secularisminstitute/> (23.06.09).
Kosmin, B. A. & Keysar, A. (2009). ARIS 2008. American religious
identification survey. Summary report, March. <http://
commons.trincoll.edu/aris/publications/aris-2008-summary-report/>.
Larson, E. J., & Witham, L. (1998). Leading scientists still reject god. Nature,
394, 313.
Lawson, E. T. (2000). Towards a cognitive science of religion. Numen, 47,
338–349.
Lewis, G. J., Ritchie, S. J., & Bates, T. C. (2011). The relationship between
intelligence and multiple domains of religious belief: Evidence from a
large adult US sample. Intelligence, 39, 468–472.
Lindeman, M., & Aarnio, K. (2006). Paranormal beliefs: Their
dimensionality and correlates. European Journal of Personality, 20,
585–602.
Lindeman, M., & Aarnio, K. (2007). Superstitious, magical, and paranormal
beliefs: An integrative model. Journal of Research in Personality, 41,
731–744.
Lynn, R., Harvey, J., & Nyborg, H. (2009). Average intelligence predicts
religiosity across 137 countries. Intelligence, 37, 11–15.
McFarland, M. J., Wright, B. R. E., & Weakliem, D. L. (2011). Educational
attainment and religiosity: Exploring variation by religious tradition.
Sociology of Religion, 72, 166–188.
Morewedge, C. K., & Kahneman, D. (2010). Associative processes in
intuitive judgment. Trends in Cognitive Sciences, 14, 435–440.
National Science Foundation (2002). Science and engineering: Indicators
2002. Science and technology: Public attitudes and public
understanding. Science fiction and pseudoscience. <http://
www.nsf.gov/sbe/srs/seind02/c7/c7s5.htm>.
Norenzayan, A., Atran, S., Faulkner, J., & Schaller, M. (2006). Memory and
mystery: The cultural selection of minimally counterintuitive
narratives. Cognitive Science, 30, 531–553.
Pacini, R., & Epstein, S. (1999). The relation of rational and experiential
information processing styles to personality, basic beliefs, and the
ratio-bias phenomenon. Journal of Personality and Social Psychology,
76, 972–987.
Petersen, L. R. (1994). Education, homogamy, and religious commitment.
Journal for the Scientific Study of Religion, 33, 122–134.
Pyysiäinen, I. (2001). How religion works. Boston, MA: Brill.
Pyysiäinen, I., & Anttonen, V. (Eds.). (2002). Current approached in the
cognitive science of religion. London, UK: Continuum.
Reeve, C. L. (2009). Expanding the g-nexus: further evidence regarding the
relations among national IQ, religiosity, and national health
outcomes. Intelligence, 37, 495–505.
Rice, Tom. (2003). Believe it or not: Religious and other paranormal
beliefs in the United States. Journal for the Scientific Study of Religion,
42, 95–106.
Shenhav, A., Rand, D. G., & Greene, J. D. (2011). Divine intuition: Cognitive
style influences belief in god. Journal of Experimental Psychology:
General. doi: 10.1037/a0025391.
Sloman, S. A. (1996). The empirical case for two systems of reasoning.
Psychological Bulletin, 119, 3–22.
Smith, M. D., Foster, C. L., & Stovin, G. (1998). Intelligence and paranormal
belief: Examining the role of context. Journal of Parapsychology, 62,
65–77.
Spinoza, B. (1982). The ethics and selected letters (S. Feldman Ed.; S.
Shirley, Trans.) Indianapolis, IN: Hackett (Originally published in
1677).
Stanovich, K. E. (2004). The Robot’s rebellion: Finding meaning in the age of
Darwin. Chicago: The University of Chicago Press.
Stanovich, K. E. (2009). Is it time for a tri-process theory? Distinguishing
the reflective and algorithmic mind. In J. St. B. T. Evans & K. Frankish
(Eds.), In two minds: Dual processes and beyond (pp. 55–88). Oxford,
UK: Oxford University Press.
Stanovich, K. E., & West, R. F. (2000). Individual differences in reasoning:
Implications for the rationality debate. Behavioral and Brain Sciences,
23, 645–726.
Stanovich, K. E., & West, R. F. (2008). On the relative independence of
thinking biases and cognitive ability. Journal of Personality and Social
Psychology, 94, 672–695.
Stark, S., & Bainbridge, W. S. (1985). The future of religions: Secularization,
revivial, and cult formation. Berkeley and Los Angeles, CA: University
of California Press.
Thompson, V. A., Prowse Turner, J., & Pennycook, G. (2011). Intuition,
reason and metacognition. Cognitive Psychology, 63, 107–140.
Tobacyk, J.J. (1988). A revised paranormal belief scale. Unpublished
manuscript, Louisiana Tech University.
Tobacyk, J. J., & Milford, G. (1983). Belief in paranormal phenomena:
Assessment instrument development and implications for personality
functioning. Journal of Personality and Social Psychology, 44,
1029–1037.
Toplak, M. V., West, R. F., & Stanovich, K. E. (2011). The Cognitive
Reflection Test as a predictor of performance on heuristics-and-biases
tasks. Memory & Cognition, 39, 1275–1289.
Zelan, J. (1968). Religious apostasy, higher education, and occupational
choice. Sociology of Education, 41, 370–379.
Zuckerman, P. (2008). Society without god: What the least religious nations
can tell us about contentment. New York: New York University Press.
Zuckerman, P. (2007). Atheism: contemporary numbers and patterns. In
M. Martin (Ed.), The Cambridge companion to atheism. Cambridge:
Cambridge University Press.
346 G. Pennycook et al. / Cognition 123 (2012) 335–346
... For example, the tendency to think analytically is associated with disbelief in conspiracy theories and experimental manipulations that induce analytical thinking decrease belief in conspiracy theories (e.g., Alsuhibani et al., 2022;Swami et al., 2014). Similarly, others have shown that the tendency to think analytically is associated with reduced paranormal beliefs above and beyond the effects of sex, age, political ideology, and education (e.g., Pennycook et al., 2012). ...
... Taken together with Supplemental Studies 1-2 we show that greater analytical reasoning is associated with better discriminability, this time in a representative of the UK population in terms of age, sex, and ethnicity. These results are consistent with research suggesting that analytical reasoning enables individuals to discern fact from fiction (Alsuhibani et al., 2022;Pehlivanoglu et al., 2021;Pennycook et al., 2012;Pennycook et al., 2015;Pennycook & Rand, 2019, 2021Swami et al., 2014) and detect phishing attempts (Ackerley et al., 2022;Bayl-Smith et al., 2020;Jones et al., 2019;Yan & Gozu, 2012). ...
... Previous research has observed that individual differences related to analytical reasoning provide a resilience against sources of internet fraud (e.g., Jones et al., 2019) and that analytical reasoning promotes the rejection of misleading, potentially harmful information (Pennycook et al., 2012;Pennycook et al., 2015;Pennycook & Rand, 2019;Swami et al., 2014). The current research represents an important attempt to connect these two literatures. ...
... To take a recent example, individuals who are more disposed toward thinking analytically (as indexed by the Cognitive Reflection Test; Frederick, 2005) are less likely to fall for fake news regardless of whether it is consistent or inconsistent with their political ideology (Pennycook & Rand, 2019b). There is also evidence that analytic thinking is associated with disbelief in a variety of epistemically suspect beliefs (Pennycook et al., 2015a), such as in paranormal and religious beliefs (Pennycook, Cheyne, Seli, Koehler & Fugelsang, 2012;Pennycook, Ross, Koehler & Fugelsang, 2016;Shenhav, Rand & Greene, 2012), conspiratorial ideation (Swami, Voracek, Stieger, Tran & Furnham, 2014), anti-science beliefs (and specifically rejection of evolution) (Gervais, 2015), and pseudo-profound bullshit receptivity (Pennycook, Cheyne, Barr, Koehler & Fugelsang, 2015). In addition, there is evidence that reliance on intuition is associated with traditional moral values (Pennycook, Cheyne, Barr, Koehler & Fugelsang, 2014;Royzman, Landy & Goodwin, 2014) and conservative political ideology (Jost, 2017) (but perhaps more-so with political apathy; see Pennycook & Rand, 2019a). ...
... Paranormal belief. Participants completed a slightly revised Paranormal Belief Scale (Pennycook, Cheyne, Seli, Koehler & Fugelsang, 2012;Tobacyk, 2004) with the religious belief items excluded ( = .95). The scale consisted of 22 items sampled from 6 categories of supernatural belief (example items in parentheses): Psi ("Mind reading is possible"), Witchcraft ("Witches do exist"), Omens of luck ("Black cats can bring bad luck"), Spiritualism ("It is possible to communicate with the dead"), Extraordinary life forms ("The Loch Ness monster of Scotland exists") and Precognition ("Astrology is a way to accurately predict the future"). ...
... God Skepticism. Skepticism about God was assessed using the following question: "What sort of God, if any, do you believe in?" and presenting the following options of increasing skepticism (Pennycook et al., 2012;Pennycook, Ross, et al., 2016) The following note was also presented: Microaggressions are defined as "brief, everyday exchanges that send denigrating messages to certain individuals because of their group membership." All corrlations are significant at < .001 ...
Article
Full-text available
Does one’s stance toward evidence evaluation and belief revision have relevance for actual beliefs? We investigate the role of endorsing an actively open-minded thinking style about evidence (AOT-E) on a wide range of beliefs, values, and opinions. Participants indicated the extent to which they think beliefs (Study 1) or opinions (Studies 2 and 3) ought to change according to evidence on an 8-item scale. Across three studies with 1,692 participants from two different sources (Mechanical Turk and Lucid for Academics), we find that our short AOT-E scale correlates negatively with beliefs about topics ranging from extrasensory perception, to respect for tradition, to abortion, to God; and positively with topics ranging from anthropogenic global warming to support for free speech on college campuses. More broadly, the belief that beliefs should change according to evidence was robustly associated with political liberalism, the rejection of traditional moral values, the acceptance of science, and skepticism about religious, paranormal, and conspiratorial claims. However, we also find that AOT-E is more strongly predictive for political liberals (Democrats) than conservatives (Republicans). We conclude that socio-cognitive theories of belief (both specific and general) should take into account people’s beliefs about when and how beliefs should change – that is, meta-beliefs – but that further work is required to understand how meta-beliefs about evidence interact with political ideology.
... Cognitive reflection, the ability to override intuitive responses, has been shown to predict individual differences in one's reasoning ability and personal beliefs. For instance, individuals higher in cognitive reflection tend to have higher SAT scores , lower levels of conventional religious beliefs (Pennycook et al. 2012), greater ability to discern "pseudoprofound bullshit" (as measured by meaningless, but syntactically correct, sentences filled with buzzwords; Pennycook et al. 2015), and rational thinking (Toplak, West, & Stanovich, 2011). ...
Thesis
Full-text available
There are ongoing debates about whether the U.S. should switch from plurality voting to alternative systems (e.g., cardinal or ranked-choice voting) and debates about the relative fairness and ease of learning different systems. To address these issues, we developed the ‘Who Won the Election Task’ (WWET) in which participants were shown the results of a hypothetical election in which a group of people were voting on which candidate to hire. The WWET had participants determine elections from raw data and allowed us to calculate the degree to which participants’ choices agreed with the three voting systems. In four studies, we evaluated how participants’ preferences about voting systems, the consistency in these preferences when measured in different ways, and whether their understanding of the voting systems and individual differences predicted their voting system preferences. Additionally, we tested educational interventions, which improved participants’ understanding of the voting systems. Across all the studies, participants’ choices in the WWET were most consistent with plurality voting. However, participants tended to view ranked-choice voting as fairer than plurality. In Studies 3 and 4 participants even sometimes viewed cardinal voting as fairer than plurality. In general, we found low consistency in voting system preferences when measured in different ways. One reason this may occur is because participants struggled to comprehend the alternative voting systems and were not adequately self-assessing their own knowledge. This research has implications for persuading the public to change voting systems for elections as well as how groups should make collective decisions (e.g., hiring decisions).
... In this sense, the cognitive abilities were suggested as a relevant determinant. Several studies have evidenced a relationship between CRT scores and cognitive abilities, such as, general mental ability [16,17,[22][23][24], numerical ability [25][26][27][28][29][30], verbal ability [31][32][33][34][35], mechanical-spatial ability [35][36][37], and other cognitive variables as working memory capacity [24,25,38], executive functions [22,39], and perceptual speed [31,40]. However, CRT scores are more robust related to general mental ability and numerical ability. ...
Article
Full-text available
This paper aims to expand the literature on the determinants of the Cognitive Reflection Test scores, exploring the effects that the items sequence has on (1) Cognitive Reflection Test scores, (2) response time, (3) the relationship between Cognitive Reflection Test scores and response time, and (4) Cognitive Reflection scores, response time, and the relationship between both variables on men and women. The current study also explored the sex differences on Cognitive Reflection Test and response time according to items sequence. The results showed that manipulating the items sequence, the performance on the Cognitive Reflection Test improved significantly, but the response time were not significantly affected, although the results suggest that first items of the sequence could be working as training items. A positive relationship between Cognitive Reflection Test scores and response time was also found, except when the scores were maximized. Finally, some differences between men and women on the results were also found. The implications of these findings are discussed.
... Type 1 processes include the automatic, effortless, and intuitive thinking that we share with our evolutionary ancestors, whereas Type 2 processes include the controlled, effortful, and reflective thinking specific to humans (Kahneman, 2011). Although the assumption of the dual-process model that the two cognitive processes are independent has recently come under scrutiny (Baron, Scott, Fincher & Metz, 2015;Białek & De Neys, 2016;Klein, 2011;Pennycook, Fugelsang & Koehler, 2015;Thompson, Evans & Frankish, 2009;Trémolière & Bonnefon, 2014), it is well-established that the relative extent of reflection vs. intuition constituting a decision-making process can nevertheless strongly influence beliefs and behaviors (e.g., ideological, re-ligious, and conspirational beliefs, and economic, moral, and health behaviors; Gervais et al., 2018;Pennycook, Cheyne, Barr, Koehler & Fugelsang, 2013;Pennycook, Cheyne, Seli, Koehler & Fugelsang, 2012;Rand, 2016;Swami, Voracek, Stieger, Tran & Furnham, 2014;Yilmaz & Isler, 2019;Yilmaz & Saribay, 2017a, 2017b. ...
Article
Full-text available
Manipulations for activating reflective thinking, although regularly used in the literature, have not previously been systematically compared. There are growing concerns about the effectiveness of these methods as well as increasing demand for them. Here, we study five promising reflection manipulations using an objective performance measure — the Cognitive Reflection Test 2 (CRT-2). In our large-scale preregistered online experiment (N = 1,748), we compared a passive and an active control condition with time delay, memory recall, decision justification, debiasing training, and combination of debiasing training and decision justification. We found no evidence that online versions of the two regularly used reflection conditions — time delay and memory recall — improve cognitive performance. Instead, our study isolated two less familiar methods that can effectively and rapidly activate reflective thinking: (1) a brief debiasing training, designed to avoid common cognitive biases and increase reflection, and (2) simply asking participants to justify their decisions.
... An analytic (or reflective) cognitive style will be defined by contrast with a more intuitive style as a greater tendency to inhibit Type-1 processes to rely on Type-2 processes. As such, "an analytic cognitive style will typically involve a broader assessment of problem elements as well as an examination and critical evaluation of intuitions" (Pennycook et al., 2012). ...
Article
Full-text available
In the context of the COVID-19 pandemic, French public opinion has been divided about Pr. Didier Raoult and his hydroxychloroquine-based treatment against COVID-19. In this paper, our aim is to contribute to the understanding of this polarization of public opinion by investigating the relationship between (analytic vs. intuitive) cognitive style and trust in Didier Raoult and his treatment. Through three studies (total N after exclusion = 950), we found that a more intuitive cognitive style predicted higher trust in Didier Raoult and his treatment. Moreover, we found that Trust in Raoult was positively associated with belief that truth is political, belief in conspiracy theories, belief in pseudo-medicines and pseudo-medical and conspiratorial beliefs regarding the COVID-19 pandemic. We also found a negative association with knowledge of scientific methods and regard for scientific method over personal experience. However, higher trust in Didier Raoult was not associated with self-reported compliance with official regulations concerning the COVID-19 pandemic.
... In contrast to the null result indicating no effect of manipulation on the main dependent variables, there was a statistically significant negative relation between the tendency to think reflectively (as evidenced by the scores on CRT) and religious belief (both measures) in one of the two conditions, in line with the previous literature (Bahcekapili Gervais & Norenzayan, 2012;Gervais et al., 2018;Pennycook et al., 2012;Stagnaro, Ross, Pennycook & Rand, 2019;Yilmaz & Saribay, 2016). Interestingly, the same discrepancy between experimental and correlational findings was observed in another recent study (Yilmaz & Isler, 2019). ...
Article
Full-text available
In 2012, two independent groups simultaneously demonstrated that intuitive mindset enhances belief in God. However, there is now some mixed evidence on both the effectiveness of manipulations used in these studies and the effect of mindset manipulation on belief in God. Thus, this proposal attempted to replicate one of those experiments (Shenhav, Rand & Greene, 2012) for the first time in a high-powered experiment using an under-represented population (Turkey). In line with the intuitive belief hypothesis, a negative correlation between reflectiveness and religious belief emerged, at least in one of the experimental conditions. In contrast to that hypothesis, however, the results revealed no effect of the cognitive style manipulation on religious belief. Although a self-report measure (Faith in Intuition) provided evidence that the manipulation worked as intended, it did not influence actual performance (Cognitive Reflection Test), suggesting a demand effect problem. Overall, the results failed to provide support for the intuitive belief hypothesis in our non-WEIRD sample, despite generally following the predicted patterns, and suggest that using stronger manipulation techniques are warranted in future studies.
Article
The main aim of the study was to replicate and extend van Prooijen's findings (2017) on how education and its outcomes (cognitive complexity, subjective social standing, self‐esteem, a feeling of control and powerlessness, cognitive reflection, epistemic curiosity and scientific reasoning) predict conspiracy beliefs. In two studies (Study 1: N = 497, Mage = 49.06, SDage = 14.92; Study 2: N = 482, Mage = 47.45, SDage = 15.87), subjective socioeconomic status and cognitive reflection (Study 1) and a feeling of powerlessness and scientific reasoning (Study 2) contributed to the negative relationship between education level and belief in conspiracy theories. These results showed the connection of education to socioeconomic status, a feeling of control and analytic thinking (cognitive reflection, scientific reasoning) and their power to reduce conspiracy beliefs. This article is protected by copyright. All rights reserved.
Article
Full-text available
Those who have shaped the modern human sciences have been preoccupied with explaining the phenomena of religion and religiosity. Accounting for the absence of religious faith has never been of much concern to them. The reasons for the neglect of atheism as a phenomenon in need of explanation are twofold. First, religion, and religiosity, seem to be adhered to by the majority of humans, and thus explaining their existence and survival is called for. Second, explaining religious beliefs is urgently called for if those doing it do not share those beliefs. And this indeed has been the case. Most of the great names in the history of modern human sciences have been atheists or agnostics, and we can recall Karl Marx, Sigmund Freud, Emile Durkheim, Bronislaw Malinowski, and Max Weber. They looked at religion from the outside, and explaining why they did not believe in any of it was unnecessary. ATHEISTS AS DEVIANTS Most theories of religion assume that it stems from universal aspects of the human condition or the human mind. Classical explanations used the notions of such universal and automatic psychological processes as projection, animism, or anthropomorphism. In some more recent theoretical formulations (Boyer 2001; Atran 2002) the framework is cognitive-evolutionary and assumes that the brain is a machine operating according to rules developed through evolution. The question is that of the seeming plausibility of religious ideas for most humans. Religion uses the basic and ordinary cognitive processes that are the evolutionary endowments of every human. The belief in supernatural agents is a by-product of naturally selected cognitive mechanisms.
Article
This paper examines the role of context in the relationship between belief in the paranormal and intellectual ability. It is argued that expressed paranormal belief is dependent upon the context in which it is measured and that such context effects may account for disparities in research findings of earlier investigators. Sixty participants were divided equally into three conditions. All participants completed the Raven Advanced Progressive Matrices, Set 1, a short test used as an indicator of intellectual ability. Participants in one condition were presented with a pro-paranormal context statement that cast paranormal belief in a positive light. Participants in the second condition were presented with an anti-paranormal statement that cast paranormal belief in a negative light. Participants in the third condition acted as a control group and were therefore given no context statement. All participants then completed Tobacyk and Milford's (1983) Paranormal Belief Scale. It was predicted that paranormal belief scores would be highest in the pro-paranormal condition, and lowest in the anti-paranormal condition. This prediction was confirmed. It was further hypothesized that this context effect would be more apparent among participants of high intellectual ability, resulting in a spurious positive correlation between belief scores and intelligence scores in the pro-paranormal condition, and a spurious negative correlation in the and-paranormal condition. No correlation was expected in the control condition. This hypothesis was not confirmed. Significant negative correlations were obtained in both the pro- and anti-paranormal conditions. A nonsignificant negative correlation was obtained in the control condition. These results do not support the notion that reports of negative correlations between paranormal belief and intellectual ability are accounted for by context effects on paranormal belief scores. Implications and limitations of these findings are discussed.
Article
Determining what percentage of a given society believes in God - or doesn't - is fraught with methodological hurdles. First: low response rates; most people do not respond to surveys, and response rates of lower than 50 percent cannot be generalized to the wider society. Second: nonrandom samples. If the sample is not randomly selected - that is, every member of the given population has an equal chance of being chosen - it is nongeneralizable. Third: adverse political/cultural climates. In totalitarian countries where atheism is governmentally promulgated and risks are present for citizens viewed as disloyal, individuals will be reluctant to admit that they do believe in God. Conversely, in societies where religion is enforced by the government and risks are present for citizens viewed as nonbelievers, individuals will be reluctant to admit that they don't believe in Allah, regardless of whether anonymity is “guaranteed. ” Even in democratic societies without governmental coercion, individuals often feel that it is necessary to say that are religious, simply because such a response is socially desirable or culturally appropriate. For example, the designation “atheist ” is stigmatized in many societies; even when people directly claim to not believe in God, they still eschew the self-designation of “atheist. ”
Article
The abstract for this document is available on CSA Illumina.To view the Abstract, click the Abstract button above the document title.
Direct fabrication of metal near-net shapes from a computer model typically involves melting and solidification, which can cause high residual stresses, undesirable phases, poor microstructures, rough surface finishes, warpage, and other problems. This paper describes a new technology, still under development, that might be used to directly fabricate solid, near-fulldensity, free-form shapes of many metals, and even some composite materials, at or near room temperature without melting and solidification. In this process, tentatively called Cold Spray Direct Fabrication (CSDF), powder particles in a supersonic jet of compressed gas impact a solid surface with sufficient energy to cause plastic deformation and consolidation with the underlying material by a process thought to be analogous to explosive welding. Material deposition by cold spray methods has already been successfully demonstrated by several investigators. This paper presents results of an experimental study to investigate the effects of selected process variables on cold spray particle velocities. In addition, a key technical barrier to the CSDF concept is focusing the spray stream down to dimensions that would permit a useful level of part detail, while still providing practical build rates. This paper presents results of initial research to develop an aerodynamic lens that may provide the required particle stream focusing.
Article
Several recent studies have examined the relationships between academic careers and other "intellectualism" variables on the one hand, and religious identification on the other. This paper presents data on this topic from a study of American arts and science graduate students. Persons who no longer identify with any religion although reared in some religion are more likely to have attended elite colleges, to manifest lower morale, to think of themselves as intellectuals and political liberals and to prefer academic careers. The patterns of relationships among these variables lend support to the inference that orientation to an academic career together with a particular constellation of secular values serves, for some, as a functional alternative to religious identification.